[clang] Add clang atomic control options and attribute (PR #114841)

[Yaxun Liu via cfe-commits]

================
@@ -5400,6 +5400,151 @@ third argument, can only occur at file scope.
     a = b[i] * c[i] + e;
   }
 
+Extensions for controlling atomic code generation
+=================================================
+
+The ``[[clang::atomic]]`` statement attribute enables users to control how
+atomic operations are lowered in LLVM IR by conveying additional metadata to
+the backend. The primary goal is to allow users to specify certain options,
+like whether atomic operations may be performed on specific types of memory or
+whether to ignore denormal mode correctness in floating-point operations,
+without affecting the correctness of code that does not rely on these behaviors.
+
+In LLVM, lowering of atomic operations (e.g., ``atomicrmw``) can differ based
+on the target's capabilities. Some backends support native atomic instructions
+only for certain operation types or alignments, or only in specific memory
+regions. Likewise, floating-point atomic instructions may or may not respect
+IEEE denormal requirements. When the user is unconcerned about denormal-mode
+compliance (for performance reasons) or knows that certain atomic operations
+will not be performed on a particular type of memory, extra hints are needed to
+tell the backend how to proceed.
+
+A classic example is an architecture where floating-point atomic add does not
+fully conform to IEEE denormal-mode handling. If the user does not mind ignoring
+that aspect, they would prefer to still emit a faster hardware atomic instruction,
+rather than a fallback or CAS loop. Conversely, on certain GPUs (e.g., AMDGPU),
+memory accessed via PCIe may only support a subset of atomic operations. To ensure
+correct and efficient lowering, the compiler must know whether the user wants to
+allow atomic operations on that type of memory.
+
+The allowed atomic attribute values are now ``remote_memory``, ``fine_grained_memory``,
+and ``ignore_denormal_mode``, each optionally prefixed with ``no_``. The meanings
+are as follows:
+ 
+- ``remote_memory`` means atomic operations may be performed on remote memory.
+  Prefixing with ``no_`` (i.e. ``no_remote_memory``) indicates that atomic
+  operations should not be performed on remote memory.
+- ``fine_grained_memory`` means atomic operations may be performed on fine-grained
+  memory. Prefixing with ``no_`` (i.e. ``no_fine_grained_memory``) indicates that
+  atomic operations should not be performed on fine-grained memory.
+- ``ignore_denormal_mode`` means that atomic operations are allowed to ignore
+  correctness for denormal mode in floating-point operations, potentially improving
+  performance on architectures that handle denormals inefficiently. The negated form,
+  if specified as ``no_ignore_denormal_mode``, would enforce strict denormal mode
+  correctness.
+
+Without any atomic metadata, LLVM IR defaults to conservative settings for
+correctness: atomic operations are assumed to use remote memory, fine-grained
+memory, and enforce denormal mode correctness (i.e. the equivalent of
+``remote_memory``, ``fine_grained_memory``, and ``no_ignore_denormal_mode``).
+
+The attribute may be applied only to a compound statement and looks like:
+
+.. code-block:: c++
+
+   [[clang::atomic(remote_memory, fine_grained_memory, ignore_denormal_mode)]]
+   {
+       // Atomic instructions in this block carry extra metadata reflecting
+       // these user-specified options.
+   }
+
+You can provide one or more of these options, each optionally prefixed with
----------------
yxsamliu wrote:

Within the same atomic attribute, duplicate and conflict values are accepted and the
last value of conflicting values wins. Multiple atomic attributes are allowed
for the same compound statement and the last atomic attribute wins.

will document these and add lit tests.

https://github.com/llvm/llvm-project/pull/114841